Download Nedocromil sodium in the treatment of exercise-induced asthma: a meta-analysis C. Spooner

Copyright #ERS Journals Ltd 2000
European Respiratory Journal
ISSN 0903-1936
Eur Respir J 2000; 16: 30±37
Printed in UK ± all rights reserved
Nedocromil sodium in the treatment of exercise-induced asthma:
a meta-analysis
C. Spooner+,*, B.H. Rowe+,#, L.D. Saunders#
Nedocromil sodium in the treatment of exercise-induced asthma: a meta-analysis. C.
Spooner, B.H. Rowe, L.D. Saunders. #ERS Journals Ltd 2000.
ABSTRACT: Exercise-induced asthma (or bronchoconstriction) afflicts millions of
people worldwide. While generally self-limiting, it can hinder performance and
reduce activity levels, thus it is an important condition to diagnose and treat. The
objective of this review was to assess the prophylactic effect of a single dose of
nedocromil sodium on exercise-induced asthma.
The Cochrane Airways Group trials register, the Cochrane Controlled Trials
Register, Current Contents, reference lists of relevant articles, review articles and
textbooks were searched for randomized trials comparing a single dose of nedocromil
to placebo to prevent exercise-induced asthma in people >6 yrs of age. Authors and the
drug manufacturer were contacted for additional trials. Trial quality assessments and
data extraction were conducted independently by two reviewers. Authors were
contacted when possible.
Twenty trials were included. All were rated as having good methodological quality.
Nedocromil inhibited bronchoconstriction in all age groups. The pooled weighted
mean difference for the maximum percentage fall in forced expiratory volume in one
second was 15.6%, (95% confidence interval (95% CI): 13.2±18.1) and for the peak
expiratory flow was 15.0% (95% CI: 8.3±21.6). These differences are both statistically
and clinically significant. After nedocromil the time to recover normal lung function
was <10 min compared to >30 min with placebo. Nedocromil had a greater effect on
people with a fall in lung function of >30% from baseline. There were no significant
adverse effects reported with this short-term use.
In conclusion, Nedocromil taken before exercise appears to reduce the severity and
duration of exercise-induced bronchoconstriction. This effect appears to be more
pronounced as severity increases.
Eur Respir J 2000; 16: 30±37.
Exercise-induced asthma (EIA) is a transient increase in
airway resistance resulting from bronchoconstriction that
occurs following 6±8 min of strenuous exercise [1]. A
postexercise fall of $10% in either the forced expiratory
volume in one second (FEV1), or the peak expiratory flow
rate (PEFR) compared with pre-exercise baselines, is considered diagnostic [2, 3]. An episode of EIA is associated
with various symptoms e.g. cough, wheeze, shortness of
breath, chest tightness, even stomach pain and nausea. It
can result in limited endurance during, and prolonged
recovery time following, the exercise. Generally, the bronchoconstriction peaks within 3±15 min post-exercise, then
spontaneously returns to pre-exercise levels within 20±60
min [4].
The true prevalence of EIA is difficult to estimate. While
it may exist as a single entity, 60±90% of people with asthma experience EIA and consider exercise a major trigger of
their asthma symptoms [1, 5]. Furthermore, it is not uncommon for asthmatics to be adequately managed and
enjoy normal lung function at rest, yet still suffer significant EIA on vigorous exercise [6]. Besides asthmatics,
35±40% of those with allergic rhinitis [7] and 12±15% of
the general population also suffer from the disorder [8].
EIA is the most common respiratory problem seen in
recreational and competitive sports (3±14%) but unfortunately, is regularly under-diagnosed and untreated and
+
Division of Emergency Medicine, University of Alberta, Edmonton, Alberta,
Canada. #Dept of Public Health Sciences
and *Institute of Health Economics, Edmonton, Alberta, Canada.
Correspondence: C.H. Spooner, Division
of Emergency Medicine, IGI.50 Walter
Mackenzie Centre, University of Alberta,
8440±112th St. Edmonton, AB T6G 2B7,
Canada. Fax: 1780 4073314
Keywords: Asthma, bronchoconstriction,
exercise-induced, meta-analysis, nedocromil, therapeutic use
Received: August 9 1999
Accepted after revision March 20 2000
This project received partial funding from
the Institute of Health Economics, an independent research centre established by the
governments of Canada and Alberta, the
Universities of Alberta and Calgary, and
10 research-based multi-national pharmaceutical companies.
A version of this review has been
published in an electronic format in the
Cochrane Library (Version 2, 2000).
thereby has the potential to severely limit the performance
of athletes [9, 10].
The severity of an EIA episode is related to the degree of
underlying bronchial reactivity [11]. Ambient air conditions, antigens, air pollutants, or respiratory viruses can
all increase bronchial reactivity and exercising while exposed to these stimuli can provoke a severe reaction and
even anaphylaxis [5, 12, 13]. Short-acting b2-agonists,
effective in reversing the bronchoconstriction, may be
required when the FEV1 or PEFR decreases by 25±30%
[14]. A small subset of asthmatics experience a second,
less severe, late-phase reaction several hours after the
original activity [5, 12].
One goal of treatment is to prevent or at least attenuate
the EIA response so respiratory limitations do not restrict
activity choices and achievement levels. Traditionally,
short-acting b2-agonist bronchodilating agents have been
the treatment of choice. However, with better understanding of the primary role of inflammatory mediators and
inflammation in the pathophysiology of asthma, several
other drugs have been investigated for their potential to
protect against bronchoconstriction. Nedocromil sodium
(NCS) is a mast cell stabilizing agent introduced in the
1980s, reportedly effective on a single-use basis for inhibiting bronchoconstriction due to antigens, fog, cold air, sulphur dioxide, and exercise [15, 16]. If effective and safe in
NCS IN PROPHYLACTIC TREATMENT OF EIA
EIA, these properties would make it an attractive therapeutic option for active people and one that may have the
added benefit of diminishing/attenuating the late-phase
reaction.
This meta-analysis examined the available randomized
controlled clinical trial evidence on the use of NCS to prevent or attenuate EIA. To the authors' knowledge, no previous systematic reviews on this topic area have been
published.
Methods
The protocol for this study was prepared prior to searching for trials, and it was reviewed and approved by the
Editorial team of the Cochrane Airways Review Group
(ARG).
Inclusion criteria
Only randomized controlled trials were included in this
systematic review. Those trials comparing a single prophylactic dose of NCS to a placebo in asthmatic individuals
(child $6 yrs or adult $18 yrs) with objective, reproducible evidence of EIA, were considered for inclusion. EIA
was defined as a fall in either the FEV1 or the PEFR of at
least 10%. If studies had more than one drug arm, only the
comparison of NCS to placebo was included. Studies using
nasal sprays were excluded.
Outcome measures
All patient outcomes were considered, however, the two
main outcomes of interest were the effect of treatment on
the maximum percentage decrease in pulmonary function
and on function over the first 30 min of recovery postexercise.
The conventional method of quantifying the severity of
EIA is to calculate the per cent fall index, which is the
maximum reduction in lung function postexercise expressed as a per cent of the pre-exercise value. In addition, it is
useful to establish the effect of drug therapy by determining
the degree of protection from airway obstruction offered by
the active drug compared to placebo (or other drug) [6].
The following formulae are used to determine these
indices:
Maximum % fall index (FEV1 or PEFR) =
imm pre value ÿ min post value
|100;
…1†
pre value
where imm=immediate; pre=pre-exercise; and post=postexercise.
Per cent protection index =
max fall placebo …%† ÿ max fall NCS …%†
|100
max fall placebo …%†
…2†
A 12% change in FEV1 or PEFR is considered clinically
significant [17, 18]. If a drug provides a protection index
of $40±50% it is considered to be a clinically significant
improvement [6].
31
The primary outcome measures reported in this review
are: 1) the mean maximum per cent fall in FEV1 or PEFR;
2) the mean per cent fall in FEV1 at varying time points
postexercise; and 3) the mean per cent protection afforded
by NCS compared to placebo.
Search strategy
Several search strategies were used to identify relevant
trials. The Cochrane Collaboration's ARG, has developed an "Asthma and Wheez* randomized controlled trial
(RCT) Register" from comprehensive searches of EMBASE, MEDLINE, and CINAHL from 1966±1999 for
controlled clinical trails [19]. The register is regularly updated using a systematic strategy for identifying RCTs
employing the following terms: placebo* OR trial* OR
random* OR double-blind OR double blind OR singleblind OR single blind OR controlled study OR comparative study. This register has been supplemented with hand
searches for additional RCTs from 20 journals in which
respiratory care articles are commonly found. There are
no language restrictions in this register nor were any
placed on the search for the present review.
An electronic search of the ARG register was performed
using the following terms: asthma OR Wheez* AND
exercise* OR exertion* AND NCS* OR Tilade. Searches
of the Cochrane Controlled Trials Register and Current
Contents, plus hand searches of the bibliography lists of
included studies, pertinent review articles and textbooks
were performed. The manufacturer of NCS (RhoÃne-Poulenc Rorer, St-Laurent, Quebec, Canada) and the authors of
included studies were asked to identify additional published, unpublished or "in-progress" studies that might have
potential relevance.
Study selection. One reviewer, who excluded citations that
were clearly irrelevant, screened the initial searches. The
second screening, conducted by two independent reviewers
using abstracts, titles and keywords, identified potentially
relevant trial reports requiring full text review. Foreign
articles were translated when necessary. Two reviewers
then independently selected trials for inclusion using defined eligibility criteria. A priori, this review was designed
to include unpublished data that met the inclusion criteria,
but to exclude data available only in abstract form. Reviewers were not blinded to authors, journal, results, or conclusions of the individual papers. Agreement was measured at
each level of screening using kappa statistics and any disagreements were resolved by discussion.
Methodological quality assessment and data extraction
Independent reviewers subjected the included trials to
two quality assessments. Firstly, to a scale that assesses
allocation concealment (Grade A: adequate concealment,
Grade B: uncertain, Grade C: clearly inadequate) [20] and
second, to a five point scale described by JADAD et al. [21]
that evaluates the quality of randomization, blinding and
reasons for withdrawal (0 = worst, 5 = best).
Using standardized forms, two reviewers independently
extracted data. Numeric calculations and graphic extrapolations were also confirmed by a second independent
reviewer. The reviewers attempted to contact at least one
author from each included trial (via Internet "people
32
C. SPOONER ET AL
searches", searches of library databases for the most recent
address of that author, contact with other reviewers in the
ARG) to confirm data extraction or to supply additional
information about the primary research.
Statistical considerations. All included studies were small
crossover trials (sample size (n) 8±24) that reported aggregate end of trial results using paired data. In this meta-analysis, the treatment and placebo group data were analysed
as though from parallel group studies. This approach would
be expected to yield a more conservative pooled estimate of
treatment effect [22, 23].
The data were analysed using Review Manager 3.0.1
(RevMan; Update Software, Oxford, UK). Pulmonary
function measures (FEV1 and PEFR) are continuous outcomes, consequently the individual study and pooled statistics are reported as a weighted mean difference (WMD)
between the treatment effect of drug and placebo with 95%
confidence intervals (95% CI). The weights given to each
study are based on the inverse of the variance [20, 24].
Heterogeneity among pooled estimates was tested with
Chi-squared statistics using an alpha of 0.10. Sensitivity
and subgroup analyses were performed to identify sources
of heterogeneity when indicated. A random effects model
was used throughout.
When the SD for the mean change in lung function for
each treatment group was not reported, an estimate was
inputed. The estimate was based on the weighted average
(using sample size) of the SDs from other included studies
for that category (i.e. adults, children, FEV1, PEFR) using
the formula [25]:
r
Pooled SD ~ …n1 ÿ 1†var1 z…n2 ÿ 1†var2
z:::z…nk 2 ÿ 1†vark
P
…3†
nÿk
where n=sample size, var=the variance of the study group
in study i, k=the number of studies with the variance
provided.
A sensitivity analysis was performed to check the effect
of inputation. Since no difference was identified, the inputation was maintained throughout the analysis.
A priori subgroup analyses were planned based on six
factors that could potentially influence the magnitude of
treatment response: 1) age (6±17 yrs were considered children, $18 yrs adults); 2) sex; 3) severity of EIA (a mean
maximum per cent fall in pulmonary function after placebo
of <30% was considered mild EIA, of $30% was considered moderate/severe EIA) [14]; 4) dose of NCS given
(#2 mg, 4 mg, $6 mg); 5) delivery system (nebulized,
metered dose inhaler (MDI), MDI with spacer); and 6)
timing of pretreatment (<30 min, $30 min).
Time course analysis
Thirteen publications [26±38] included graphs that
compared lung function post-treatment at various times
up to 120 min postexercise. Unfortunately, the absolute
values and SDs to accompany the graphs were not reported. To extract this valuable data, the graphs were enlarged and two reviewers visually estimated the mean per
cent fall FEV1 values for the two groups at each time
point. Where estimates between reviewers differed, the
average of the two was used. The mean and SD for each
time point was then calculated. The Mann-Whitney U-test
was used to assess the significance of the differences at
each time point using an alpha of 0.05.
Results
Search
Fifty-one trials, including two German, one Spanish and
one Italian, were screened for potential inclusion. A total of
20 published [26±45] and one unpublished study [46] met
the inclusion criteria however, none were foreign language trials. SINCLAIR and WINFIELD [37] did not report
mean maximum per cent fall indices but did report data
that were included in the time course results. Reasons for
exclusion included "participants had mixed or uncertain
diagnoses" (n=1), "selection of known responders" (n=2),
"nonrandom allocation" (n=2) "abstract publication only"
(n=1), "active drug comparisons" (n=2), "not selected
outcomes" (n=2), "not primary research" (n=19).
In total, 20 studies are included in the meta-analysis and
13 studies in the time-course analysis. Nine of the 15
(60%) primary authors were successfully contacted. Where
appropriate, authors confirmed that subjects participated in
only one trial. Any duplicate data were included only once
in the analyses.
Methodological quality
All trials included were rated as being of good to high
quality using either the COCHRANE [20] or JADAD et al. [21]
criteria. (table 1). All studies were randomized doubleblind, and either described drop-outs or had none. Three
studies were rated as having "adequate" concealment of
allocation [28, 40, 45] all others rated an "unclear" status.
(kappa = 0.88). Using the five-point validity scale of
JADAD et al. [21], two studies were rated five (strong),
nine were rated four (very good), and 10 were rated three
(good) (table 1). Agreement on quality scoring was good
(kappa = 0.67).
Description of studies
The analysis includes data on 280 participants: (64%/
36% male/female); 58% were children aged 6±17 yrs, 42%
were adults aged 18±59 yrs. No females who were either
pregnant or possibly pregnant were included in any studies. Two studies selected only nonsmokers, otherwise smoking history was not clear. All subjects were described as
being "stable asthmatics" at the time of challenge testing.
TODARO [38] studied Olympic level athletes, none of the
remaining trialists addressed level of physical conditioning.
The studies compared NCS to placebo and in some
cases to one other agent (sodium cromoglycate (SCG) n=8
[29, 31, 32, 37, 41±44], furosemide n=1 [33], minocromil
n=1 [35]). Concurrent therapy included a variety of
common anti-asthma agents but most medications were
discontinued for periods of 6 h to 1 week prior to each
challenge to effect a washout and to limit confounding
influences.
Table 1. ± Characteristics of included studies
Country
Time
between
challenges
Index for
diagnosis
% fall
Drug dose
compared to
placebo
4
Italy
4 days
15%
4
Italy
2 days
3
Denmark
5
Jadad
Quality
score
BONER [26]
BONER [27]
Author [Ref.]
BUNDEGAARD [39]+
CHUDRY [40]
#
COMIS [41]+
Delivery
system
Pre-exercise
therapy min
Sample
size
Age
yrs
Sex
M/F
PFT
reported
4 mg NCS
MDI, MDI
+AA spacer
15
13
10 (7.5±13)
9/4
FEV1
15%
4 mg NCS
MDI
30
20
11.3 (7.5±15)
15/5
FEV1, PEFR,
FVC, FEF25±75
2 wks
20%
2 mg NCS 4 mg NCS
MDI
30
14
31 (21±49)
6/8
PEFR
UK
2 days in 1 wk
20%
4 mg NCS
12
13.9 (8±15)
9/3
FEV1
12
11 (6.5±13.5)
7/5
FEV1
MDI
30, 150, 270
MDI &
MDI+spacer
30
Italy
daily
15%
4 mg NCS, 10 mg SCG
DEBENEDICTUS [29]
3
Italy
separate days
15%
4 mg NCS, 10 mg SCG
MDI
20
17
10.2 (7±15)
11/6
FEV1
DEBENEDICTUS [42]
+
3
Italy
separate days
15%
4 mg NCS, 10 mg SCG
MDI+spacer
20
8
8.7 (7±11)
5/3
FEV1
DEBENEDICTUS [43]
+
3
Italy
separate days
15%
4 mg NCS, 10 mg SCG
MDI
20 & 140
13
10 (7±15)
9/4
FEV1
DEBELIC [28]
3
Germany
separate days
20%
4 mg NCS
MDI
30
12
16.9 (14±19)
7/5
FEV1
HENRIKSEN [30]
4
Denmark
1 wk
20%
4 mg NCS
MDI
30
12
10.8 (7±14)
10/2
FEV1 PEFR
3
USA
separate days
20%
4 mg NCS, 20 mg SCG
MDI
20, 120, 240
12
27.3 (21±38)
12/0
FEV1
3
Australia
separate days
20%
4 mg NCS
MDI
15
14
NR(15±45)
NR/NR
PEFR
#,+
KONIG [44]
MIHALYKA [46]
+
MORTON [31]
4
Australia
separate days
15%
8 mg NCS 4 mg SCG
MDI
15
16
20 (13±30)
10/6
FEV1
NOVEMBRE [32]
3
Italy
separate days
15%
4 mg NCS 10 mg SCG
MDI+spacer
20 min
19
NR (6±15)
13/6
FEV1, PEFR,
FEF25±75%
NOVEMBRE [33]+
3
Italy
separate days
15%
4 mg NCS 30 mg
furosemide via nebulizer
MDI+spacer
20 min
24
NR (6±16)
16/6
FEV1, PEFR,
FEF25±75%
OSEID [34]
4
Norway
separate days
20%
4 mg NCS
MDI
30 min
9
31.1 (16±50)
6
FEV1, FVC,
FEF25±75%
ROBERTS [35]1
4
UK
separate days
20%
2 & 4 mg NCS 4 mg
minocromil
MDI
30 min
9
31.1 (16±50)
6
FEV1
SHAW [36]
4
UK
1 wk
20%
2 mg NCS
MDI
20 min
8
NR (17±47)
8
FEV1, FVC
SINCLAIR [37]
5
UK
separate days
15%
4 mg NCS 10 mg SCG
MDI
30 min
20
20.7 (17±28)
18
FEV1
TODARO [38]
3
Italy
separate days
15%
4 mg NCS
MDI
20 min
13
25 (19±31)
11
FEV1
Norway
separate days
20%
1 mg, 4 mg, & 8 mg NCS
MDI
60 min
12
29 (20±45)
9
PEFR
+
1
VILSVIK [45]
4
+
#
NCS IN PROPHYLACTIC TREATMENT OF EIA
3
+
1
33
*: mean (range); : 3-arm study; : duration study; : 4-arm study. PFT: pulmonary function test; NCS: nedocromil sodium; MDI: metered dose inhaler; AA spacer: Auty-Altounyan spacer
device; FEV1: forced expiratory volume in one second; PEFR: peak expiratory flow rate; FVC: forced vital capacity; FEF25±75: forced mid-expiratory flow; SCG: sodium cromoglycate; NR:
not reported.
34
N
P
Fixed estimate
Random effects
-100
-50
0
WMD %
50
100
Fig. 1. ± The weighted mean difference (WMD, X) in the maximum per
cent fall in forced expiratory volume in one second (FEV1) between
treatment with nedocromil (N) and placebo (P). The horizontal bars
represent the 95% confidence interval. Values either side of 0% favour
either N or P, as shown. PE: pooled estimate.
Study
N
P
-100
-50
0
WMD %
50
100
Fig. 2. ± The weighted mean difference (WMD, X) in the maximum per
cent fall in peak expiratory flow rate (PEFR) between treatment with
nedocromil (N) and placebo (P). The horizontal bars represent the 95%
confidence interval. Values either side of 0% favour either N or P, as
shown. PE: pooled estimate.
geneity in this pooled result was demonstrated (Chi-squared = 20.28, df = 6, p<0.001). Neither subgroup comparisons based on the age, dose, delivery system, or timing
explained the heterogeneity, nor sensitivity analysis based
on study quality (JADAD et al. [21] score #3 versus scores
$4), imputed standard deviations (n=2), random versus
fixed effects models or publication status.
Exercise-induced asthma severity subgroup
Heterogeneity was clinically and statistically insignificant when studies were dichotomized into mild versus
moderate/severe EIA groups (fig. 3). This analysis indicated that NCS inhibited the mean maximum per cent fall
PEFR in subjects with moderate/severe EIA, to a significantly greater degree than in subjects with mild EIA:
WMD = 25.1% (95% CI: 18.7±31.1) compared to WMD
= 8.3% (95% CI: 4.1±12.5) respectively. Similar results
were demonstrated for mean maximum per cent fall
FEV1, WMD = 21.4% (95% CI: 17.2±25.5) for moderate/
severe EIA compared to WMD = 12.8% (95% CI: 10.0±
15.7) for mild EIA (fig. 4).
BONER [27]
MIHALYKA [46]
NOVEMBRE [33]
NOVEMBRE [32]
Pooled estimate
BUNDGAARD [39]
HENRIKSEN [30]
VILSVIK [45]
Pooled estimate
Study
BONER [26]
BONER [27]
CHUDRY [40]
COMIS [41]
DEBENEDICTIS [29]
DEBENEDICTIS [42]
DEBENEDICTIS [43]
DEBELIC [28]
HENRIKSEN [30]
KONIG [44]
MORTON [31]
NOVEMBRE [33]
NOVEMBRE [32]
OSEID [34]
ROBERTS [35]
SHAW [36]
TODARO [38]
Random
effects model
Fixed effects model
Mild EIA
PE
Study
Pulmonary function results. All trials reported comparable
outcome data that originated from populations and
interventions that were similar in nature; consequently,
the reviewers felt it was appropriate to combine the results
for a quantitative pooled estimate of treatment effect.
Following NCS therapy, there was significant attenuation of the mean maximum per cent fall FEV1 and PEFR
postexercise when compared to placebo therapy. The effect
size did not differ significantly in further subgroup analysis
based on age, dose, delivery system, or timing of pretreatment. Seventeen trials reporting changes in FEV1 were
pooled (fig. 1). The mean maximum per cent fall FEV1 for
placebo = 31.9% (95% CI: 28.1±35.8), for NCS = 15.6%
(95% CI: 13.4±17.8). The WMD between the two =
15.6% (95% CI: 13.2±18.1). The Chi-squared test for heterogeneity was nonsignificant for this pooled result (Chisquared = 17.57, degrees of freedom (df) = 16, p>0.25).
Seven trials reporting changes in PEFR were also
pooled (fig. 2). The positive effect on the mean maximum
per cent fall PEFR was of a similar magnitude. The mean
maximum per cent fall PEFR for placebo = 30.3% (95%
CI: 21.0±39.6), for NCS = 14.5% (95% CI: 11.3±17.7).
WMD = 15.0% (95% CI: 8.3±21.6). Significant hetero-
BONER [27]
BUNGAARD [39]
HENRIKSEN [30]
MIHALYKA [46]
NOVEMBRE [33]
NOVEMBRE [32]
VILSVIK [45]
Severe EIA
No data on symptom scores, performance measures,
patient preference, or physiological measures were reported; consequently, pooled analyses were not possible.
Twelve of 20 studies (60%) commented on adverse effects.
Seven of these (58%) stated that no adverse effects were
noticed in the time period that participants were observed.
Five studies mentioned minor side effects that included a
bad taste, throat irritation, and cough. One study reported a
mean increase in heart rate after NCS; however, the difference was not clinically significant (4 beats.min-1).
Higher doses of NCS did not appear to increase the side
effect profile.
PE
C. SPOONER ET AL
-100
N
P
-50
0
WMD %
50
100
Fig. 3. ± The weighted mean difference (WMD, X) in the maximum per
cent fall peak expiratory flow rate (PEFR) between nedocromil (N) and
placebo (P) by severity of exercise-induced asthma (EIA). The horizontal bars represent the 95% confidence interval. Values either side of
0% favour either N or P, as shown.
35
P
Study
120–150 min
postexercise
N
240–270 min
postexercise
BONER [26]
BONER [27]
DEBENEDICTIS [29]
DEBENEDICTIS [43]
KONIG [44]
NOVEMBRE [33]
NOVEMBRE [32]
OOSEID [34]
TODARO [38]
Pooled estimate
-100
CHUNDRY [40]
DEBENEDICTIS [43]
KONIG [44]
Pooled estimate
N
P
CHUDRY [40]
KONIG [44]
Pooled estimate
-100
CHUDRY [40]
COMIS [41]
DEBENEDICTIS [42]
DEBELIC [28]
HENRIKSEN [30]
MORTON [31]
ROBERTS [35]
SHAW [36]
Pooled estimate
-50
0
WMD %
50
100
Fig. 5. ± The weighted mean difference (WMD, X) in the maximum per
cent fall in forced expiratory volume in one second (FEV1) between
nedocromil (N) and placebo (P): duration. The horizontal bars represent
the 95% confidence interval. Values either side of 0% favour either N or
P, as shown.
Discussion
-50
0
WMD %
50
100
Fig. 4. ± The weighted mean difference (WMD, X) in the maximum per
cent fall in forced expiratory volume in one second (FEV1) between
nedocromil (N) and placebo (P) by severity of exercise-induced asthma
(EIA). The horizontal bars represent the 95% confidence interval. Values
either side of 0% favour either N or P, as shown.
The protection index
On average, NCS provided a clinically significant protection index of 51% (95% CI: 49±52) against a decrease
in FEV1 (range 30±70%) corroborated by a protection
index of 49% (95% CI: 47±51) against a decrease in PEFR
(range 33±66%).
Dosage sensitivity analyses
Although no definite trend was noted when comparing
low, moderate or high doses of NCS, note that few trials
investigated a low (#2 mg, n=3) or a high (>4 mg, n=1)
dose and this prevented any meaningful analyses for a
dose-response effect. The recommended clinical dose of 4
mg was used in 16/20 studies.
Duration of effect. Three trials [40, 43, 44] recorded the
mean maximum per cent fall FEV1 on 2 or 3 repeated
challenges to evaluate NCS for its duration of effect (fig. 5).
No significant benefit of NCS over placebo was demonstrated beyond the first challenge. WMD = 5.9% (95% CI:
1.0±12.9; n=3), and WMD = 5.7% (95% CI: 2.8±14.2;
n=2) at 2±2.5 h and 4±4.5 h respectively.
Time course analysis. In the 13 studies that reported the
mean per cent fall in FEV1 at seven time points during the
immediate postexercise period following both therapies,
statistically significant improvements in FEV1 measures
favouring NCS were noted at all times up to 30 min
postexercise (p<0.001) (fig. 6). The return to normal lung
function (i.e. <10% change from baseline) occurred within 10 min postexercise using NCS compared to 30 min
using placebo.
This large systematic review is the first of its kind to
examine the efficacy of NCS in preventing the bronchoconstriction typically observed in EIA. The results of this
review clearly document the benefit of NCS in attenuating
the maximum per cent fall in pulmonary function, blunting
the entire EIA response, and identifying a more rapid return to baseline lung function. This effect is consistent
across a wide variety of studies regardless of the subgroup
and sensitivity comparisons performed. Furthermore, this
attenuation is both statistically and clinically important.
There does not appear to be a dose response curve
associated with NCS delivery, however, there were insufficient data for a meaningful comparison. The degree of
protection is most impressive in those patients with more
severe EIA. The drug appears to be well tolerated, is simple to use, and may be delivered by MDI with or without a
spacer device.
Clearly, the statistically and clinically important effects
of NCS indicate that it should be considered for use in EIA.
While other agents are available to treat the disorder, the
relative merits of one drug compared to another are not
well established. From a clinical perspective, NCS was
Mean maximum fall FEV1 %
Severe EIA
Study
Mild EIA
NCS IN PROPHYLACTIC TREATMENT OF EIA
5.00
0.00
-5.00
-10.00
-15.00 ■
-20.00
-25.00
-30.00
-35.00
-40.00
-45.00
0–1
■
■
■
■
■
■
3
5
10
15
Time postexercise min
20
30
Fig. 6. ± Treatment effect of nedocromil and placebo on the time-course
of exercise-induced asthma. Data are mean maximum fall in forced
expiratory volume in one second (FEV1) and 95% confidence interval.
&: placebo; X: nedocromil.
36
C. SPOONER ET AL
well tolerated and adverse effects were minor, therefore, it
may be useful for patients who dislike the autonomic sideeffects of b2-agonists or the adverse effects of other
choices. It may also have a role in combination therapy for
people who experience EIA symptoms despite premedication with another drug [47].
There are a variety of issues that must be addressed to
produce a valid systematic review or meta-analysis. Most
importantly, systemic reviews must address the issues of
study publication and selection bias. This review employed
comprehensive search strategies that identified published,
unpublished, and foreign language literature, in addition to
employing an unbiased selection process to decide on inclusion. There remains the possibility of publication bias
due to having missed retrieving unpublished negative
trials. If this is the case, the pooled effect of NCS therapy
may be overestimated.
Comparisons in trial design, populations, interventions
studied, and methods of outcome reporting are also important issues in systemic review methodology. In this review,
only randomized controlled trials were included, ensuring
the highest level of evidence was incorporated. Studies
were pooled only after similarities in populations, interventions and outcomes were determined, and pooling was
accomplished using accepted mathematical approaches.
Despite the methodological care taken, some limitations
to generalizability remain. To assess the efficacy of NCS it
was necessary for the exercise challenges to take place in
laboratories with controlled environments. To assess the
effectiveness of the drug under normal conditions, the drug
needs to be evaluated by asthmatics during regular exercise
where environmental conditions have greater variability.
To control for variations in baseline severity of EIA on
the magnitude of drug effect, the reviewers selected the
mean maximum per cent fall index on the placebo challenge for comparison. This was the only subgroup analysis
that demonstrated a significant difference in effect size. In
planning a new primary study, stratifying by this variable
may reveal additional information on who could best benefit from pretreatment with NCS. Studying an increasing
dose-effect in nonresponders and known responders could
also reveal important information.
All studies in this review used the crossover design.
Future studies using the crossover method must concentrate on complete reporting of results by period and sequence to assure that readers' concerns over "carry-over"
and "period effects" have been addressed. The reviewers
believed the potential for a carry-over effect to be negligible for two reasons. Firstly, NCS is a short acting agent
(half-life 1.5±2 h) with topical effects. While the drug is
well absorbed from the lung, it does not accumulate systemically, has no known systemic effects, is rapidly cleared
from the body, and has few side-effects [15, 16, 48]. Secondly, the exercise challenges were conducted on separate
days, thus the trialists adhered to the recommended washout time of 5±10 times the half-life of a drug [49, 50] in the
unlikely event of systemic effects. Had there been a
period effect in every study, there is no reason to believe
that a systematic bias favours any one challenge period.
The large number of studies included should ensure an
equal distribution of period effects if they exist. By averaging the estimates, any effect would disappear leaving
an unbiased estimate of the treatment contrast [51].
Finally, data on symptom scores, exercise performance,
or subject satisfaction were not included in the studies. The
patient's own assessment of NCS is clearly an important
consideration in choosing one treatment over another.
In conclusion, this review of 20 randomized trials with
280 adults and children across eight countries over 10 yrs,
supports the single dose use of nedocromil as an effective
pharmaceutical option for the management of exerciseinduced asthma. When nedocromil was taken within an
hour of an intense, prolonged exercise challenge, the severity of bronchoconstriction, measured by the change in forced expiratory volume in one second or peak expiratory
flow rate, was significantly reduced. There was evidence to
indicate that the exercise-induced asthma response was
blunted over the entire immediate postexercise period and
that, on average, people returned to normal lung function
within 10 min of completing the exercise challenge. Considering the methodological rigour, consistency of findings
over subgroup and sensitivity analyses, the reviewers feel
that the results of this review are valid. Moreover, they
should be generalizable to a broad range of patients with
exercise-induced asthma.
Acknowledgements. The reviewers wish to
acknowledge the assistance provided by the
Cochrane Airways Review Group: P.W. Jones
(Co-ordinating editor), and S. Milan, A. Bara,
and J. Dennis (staff). The reviewers are grateful
to J. Spooner, I. Wenger, and A. Travers, I. GamezNava, L. Gonzalez-Lopez, A. Vigano and K.
Froese for their help with data abstraction and
interpretation of foreign language literature. The
reviewers would like to thank the following authors
who responded and provided additional data or
clarification when possible: S. Anderson, A. Boner,
F. De Benedictis, M. Debelic, H. Magnussen, R.
Shaw, M. Silverman, D. Sinclair, and A. Todaro.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Anderson SD. Issues in exercise-induced asthma. J
Allergy Clin Immunol 1985; 76: 763±772.
Anderson SD, Silverman M, Konig P, Godfrey S.
Exercise-induced asthma. Br J Dis Chest 1975; 69: 1±39.
Kattan M, Keens TG, Mellis CM, Levison H. The
response to exercise in normal and asthmatic children. J
Pediatr 1978; 92: 718±721.
Virant FS. Exercise-induced bronchospasm: epidemiology, pathophysiology, and therapy. Med Sci Sports Exerc
1992; 24: 851±855.
Rupp NT. Diagnosis and management of exercise-induced asthma. Phys Sportsmed 1996; 24: 77±87.
Anderson SD. Specific problems: exercise-induced asthma. In: O'Byrne P, Thomson NC, eds. Manual of Asthma
Management. London, UK, WB Saunders Co. Ltd, 1995;
pp. 621±643.
Nastasi KJ, Heinly TL, Blaiss MS. Exercise-induced
asthma and the athlete. J Asthma 1995; 32: 249±257.
Spector SL. Update on exercise-induced asthma. Ann
Allergy 1993; 71: 571±577.
Hogshead N, Couzens GS. Asthma and Exercise. New
York, NY, USA, Henry Holt and Company, 1989; pp. 1±
114.
Pierson WE, Voy RO. Exercise-induced bronchospasm in
the XXIII summer Olympic Games. N Engl Reg Allergy
Proc 1988; 9: 209±213.
NCS IN PROPHYLACTIC TREATMENT OF EIA
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
McFadden ER, Gilbert IA. Exercise-induced asthma. N
Engl J Med 1994; 330: 1362±1367.
Hendrickson CD, Lynch JM, Gleeson K. Exercise induced asthma: a clinical perspective. Lung 1993; 172: l±
14.
Du Buske LM, Horan RF, Sheffer AL. Exercise-induced
allergy syndromes. In: Weiler JM, ed. Allergic and
Respiratory Disease in Sports Medicine. New York, NY,
USA, Marcel Dekker, Inc., 1997; pp. 253±278.
Anderson SD, Daviskas E. Pathophysiology of exerciseinduced asthma: the role of respiratory water loss in
allergic and respiratory disease. In: Weiler JM, ed. Allergic and Respiratory Disease in sports Medicine. New
York, NY, USA, Marcel Dekker, Inc., 1997; pp. 87±114.
Holgate ST. Clinical evaluation of nedocromil sodium in
asthma. Eur J Respir Dis 1986; 69: Suppl. 147, 149±159.
Auty RM. The Clinical development of a new agent for
the treatment of airway inflammation, nedocromil sodium
(Tilade1). Eur J Respir Dis 1986; 69: Suppl. 147, 120±
131.
Canadian Thoracic Society. Canadian Asthma Consensus
Conference: Summary of recommendations. Can Respir
J 1996; 3: 89±100.
American Thoracic Society. Guidelines for the evaluation
of impairment/disability in patients with asthma. Am Rev
Respir Dis 1993; 147: 1056±1061.
Bara A, Milan S, Jones PW. Identifying asthma RCTs
with Medline and Embase. 3rd Annual Cochrane Colloquium. Oslo, Norway, 1995.
The Cochrane Collaboration. Cochrane Collaboration
Handbook. The Cochrane Library [database on disk and
CDROM]. updated Mar 1997. Oxford, UK, Update Software, 1997.
Jadad AR, Moore RA, Carroll D, et al. Assessing the
quality of reports of randomised clinical trials: is blinding
necessary? Control Clin Trials 1996; 17: 1±12.
Senn S. The AB/BA crossover: past, present and future?
Stat Methods Med Res 1994; 3: 303±324.
Kenward MG, Jones B. The analysis of data from 2 6 2
crossover trials with baseline measurements. Stat Med
1987; 6: 911±926.
Petitti DB. Meta-analysis, Decision analysis, and costeffectiveness analysis: methods for quantitative synthesis
in medicine. New York, NY, USA, Oxford University
Press, 1994; pp. 90±110.
Follman D, Elliott P, Suh I, Cuter J. Variance imputation
for overviews of clinical trials with continuous response.
J Clin Epidemiol 1992; 45: 769±773.
Boner AL, Miglioranzi P, Piacentini GL, Peroni DG,
Andreoli A. Effects of nedocromil sodium pressurized
aerosol on exercise challenge using a spacer device and
the normal adapter. Pediatr Asthma Allergy Immunol
1988; 2: 207±213.
Boner AL, Vallone G, Bennati D. Nedocromil sodium in
exercise-induced bronchoconstriction in children. Ann
Allergy 1989; 62: 38±41.
Debelic M. Nedocromil sodium and exercise-induced
asthma in adolescents. Eur J Respir Dis 1986; 69 (Suppl
147): 266±267.
DeBenedictis FM, Tuteri G, Bertotto A, Bruni L, Vaccaro
R. Comparison of the protective effects of cromolyn sodium and nedocromil sodium in the treatment of exerciseinduced asthma in children. J Allergy Clin lmmunol 1994;
94: 684±688.
Henriksen JM. Effect of nedocromil sodium on exerciseinduced bronchoconstriction in children. Allergy 1988;
43: 449±453.
Morton AR, Ogle SL, Fitch KD. Effects of nedocromil
sodium, cromolyn sodium, and a placebo in exerciseinduced asthma. Ann Allergy 1992; 68: 143±148.
Novembre E, Frongia GF, Veneruso G, Vierucci A. Inhibition of exercise-induced-asthma (EIA) by nedocromil
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
37
sodium and sodium cromoglycate in children. Pediatr
Allergy Immunol 1994; 5: 107±110.
Novembre E, Frongia GF, Lombardi E, Veneruso G,
Vierucci A. The preventive effect of nedocromil or furosemide alone or in combination on exercise-induced asthma in children. J Allergy Clin Immunol 1994; 94: 201±
206.
Oseid S, Mellbye E, Hem E. Effect of nedocromil sodium
on exercise-induced bronchoconstriction exacerbated by
inhalation of cold air. Scan J Med Sci Sports 1995; 5: 88±
93.
Roberts JA, Thomson NC. Attenuation of exercise-induced asthma by pretreatment with nedocromil sodium
and minocromil. Clin Allergy 1985; 15: 377±381.
Shaw RJ, Kay AB. Nedocromil, a mucosal and
connective tissue mast cell stabilizer, inhibits exerciseinduced asthma. Br J Dis Chest 1985; 79: 385±389.
Sinclair DG, Winfield CR. Attenuation of exercise-induced asthma by nedocromil sodium and sodium cromoglycate. J Royal Army Med Corps 1990; 136: 105±106.
Todaro A, Faina M, Alippi B, Dal Monte A, Ruggieri F.
Nedocromil sodium in the prevention of exercise-induced
bronchospasm in athletes with asthma. J Sports Med
Physl Fitness 1993; 33: 137±145.
Bundgaard A, Enehjelm SD, Schmidt A. A comparative
study of the effects of two different doses of nedocromil
sodium and placebo given by pressurised aerosol in exercise-induced bronchoconstriction. Eur J Allergy Clin
Immunol 1988; 43: 493±496.
Chudry N, Correa F, Silverman M. Nedocromil sodium
and exercise induced asthma. Arch Dis Child 1987; 62:
412±414.
Comis A, Valletta EA, Sette L, Andreoli A, Boner AL.
Comparison of nedocromil sodium and sodium cromoglycate administered by pressurised aerosol, with and
without a spacer device in exercise- induced asthma in
children. Eur Respir J 1993; 6: 523±526.
DeBenedictis FM, Tuteri G, Niccoli A, Mezzetti D, Rossi
L, Bruni L. The effect of cromolyn sodium and nedocromil sodium administered by a pressurized aerosol with
a spacer device on exercise-induced asthma in children.
Mediators Inflamm 1994; 3 Suppl. 1, S35±S37.
DeBenedictis FM, Tuteri G, Pazzelli P, Bertotto A, Bruni
L, Vaccaro R. Cromolyn versus nedocromil: duration of
action in exercise-induced asthma in children. J Allergy
Clin Immunol 1995; 96: 510±514.
Konig P, Hordvik NL, Kreutz C. The preventive effect
and duration of action of nedocromil sodium and cromolyn sodium on exercise-induced asthma (EIA) in adults. J
Allergy Clin Immunol 1987; 79: 64±68.
Vilsvik J, Schaanning J. A comparative study of the effect
of three doses of nedocromil sodium and placebo given
by pressurised aerosol to asthmatics with exercise-induced bronchoconstriction. Ann Allergy 1988; 61: 67±70.
Mihalyka MS, Anderson SD, Corte P. Nedocromil sodium in exercise induced asthma. Aust NZ Med 1987; 17:
Suppl. 2, 524.
DeBenedictis FM, Tuteri G, Pazzelli P, Solinas LF,
Niccoli A, Parente C. Combination drug therapy for the
prevention of exercise-induced bronchoconstriction in
children. Ann Allergy Asthma lmmunol 1998; 80: 352±
356.
Canadian Pharmaceutical Association. Compendium of
Pharmaceuticals and Specialties. 32 ed. Ottawa, Canada,
Canadian Pharmaceutical Association, 1997.
Lloyd J. Clinical trial design. In: Lloyd J, Raven A, eds.
Handbook of Clinical Research 2 ed. Edinburgh, UK,
Churchill Livingstone, 1994; pp. 92±113.
Shapiro SH, Louis TA. Clinical Trials: Issues and Approaches. New York, NY, USA, Marcel Dekker Inc., 1983.
Senn S, Hildebrand H. Crossover trials, degrees of freedom, the carryover problem and its dual. Stat Med 1991;
10: 1361±1374.